
Date: Fri, 23 Feb 2001 13:20:28 +0100
From: Peter Kretschmar <peter.kretschmar@obs.unige.ch>
Subject: Re: parameters
To: Carol Anne Oxborrow <oxborrow@dsri.dk>
Cc: stefan@astro.su.se, smaisala@astro.helsinki.fi, allan@dsri.dk, njw@dsri.dk, juhani.huovelin@astro.helsinki.fi
MIME-version: 1.0
Content-transfer-encoding: 7bit
X-Accept-Language: de,fr

Carol Anne Oxborrow wrote:
> 
> while I agree that the 20A format for OBTs is very unwieldy for entering the
> binning start time, I'm not sure that IJD is much of an improvement if you're
> sitting with a raw file in front of you and want to do some binning. [...]
> 
  Well, you would now the IJD of the start of your Science Window in IJD,
this is encoded as TSTART, TSTOP in the SWG (and possibly copied to all
children).

  My main point is that setting a specific zero time for binning is mostly 
required when you want to align your data with some other observation
or external constraint. And in this case it would be usually easier to
enter IJD, which is after all just Julian Date - constant.

  If one just wants to exclude a part of the ScW, this should (IMHO) 
_not_ be done by timeZero, timeEnd, but by definition of appropriate
GTIs.

  Can anyone provide me with a good example, where we would need a timeZero
that is relative to the ScW? 

> In all the fits files, OBT is stored as 4 floats, [...]

4 unsigned integers, not floats!

> 
> On the other hand, we already have two uniform parameter names, startRow and
> endRow that can be used to define the first and last events in the raw data 
> to be used for event processing [...]
>
  To me this looks as akward as the other point. And in higher level
processing, I would rather have GTIs than any selection row by row.

  Regards,
  Peter
